In the past decade, populist parties have garnered unexpected support in the Western world, with recent literature pointing to the diffusion of the internet as a possible determining factor. However, the role of social media, remains less clear.

Populist politicians have increasingly utilized platforms like Twitter to gain access to mainstream media attention, potentially altering the political landscape. This study aims to explore whether the success of populist politicians on Twitter led to increased legacy media attention and, ultimately, greater electoral success.

To identify the effect of Twitter on media attention, I analyze whether greater media focus on Twitter led to more press mentions for populist politicians. To avoid endogeneity, I exploit exogenous variation in media interest towards Twitter, leveraging the early-stage boost in Twitter adoption during the 2007 SXSW Festival. As documented by Müller (2020), the festival significantly accelerated Twitter’s growth, tripling daily tweets. Measuring the number of participants to the event by their home county, I predict local media attention to Twitter, proxied by mentions of "Twitter" or "tweet" in newspapers. The hypothesis is that counties with higher twitter usage would drive greater demand for Twitter-related news, and journalists in those areas would be more likely to write about the platform and its content. Instrumenting media attention with this exogenous variation, I expect to find that Twitterrelated media attention predicts a higher number of press mentions for both populist and non-populist candidates active on Twitter. However, I presume to identify a larger effect for populist candidates, suggesting that Twitter discourse particularly benefited populists in gaining media exposure. 

A preliminary analysis is presented leveraging a dataset of over 25 million unique newspaper articles from the 100 most widely read newspapers in the U.S., spanning from 1996 to the present, and more than 100 thousands twitter post by 1300 US politicians twitter accounts. While the proof of concept results are already in line with expectations, the final project will scale up to 1,230 newspapers and at least 2000 US politicians twitter accounts for broader coverage.

Future work will also incorporate data from the Cooperative Congressional Election Study (CCES) to assess whether this increased media attention influenced political preferences and voter turnout. This extension will help clarify whether the higher press mentions of populists also impacted electoral behavior and political engagement.